Why the Open OS Model Failed in Smartphones

Fifteen years ago, when Microsoft ruled the world and Apple was near death, the tech world was convinced that the conceptual batter between Windows and Mac–open operating systems available to all comers vs. closed systems–had been decided firmly in favor of open. But what applied to PCs in the 1990s does not appear to work at all for smartphones in the 2010s, as Google’s planned purchase of Motorola Mobility marks the beginning of the end for the open OS approach.

A major reason for this is that phones–and tablets–are very different from PCs even though they perform many of the same functions. A phone is a much more tightly integrated device in which it is very difficult to tell where the hardware ends and the software begins. Getting the user experience just right is both harder and more critical, because quirks that are a minor annoyance on a PC–or which can be remedied through an accessory such as a better mouse or keyboard–become killer flaws.

It’s easy to forget today that the first real winner in the smartphone market was Research In Motion’s BlackBerry, a closed system. RIM’s accomplishment was to provide a tightly controlled, secure mobile email device (the earliest models offered neither voice not internet service) that provided seamless access to corporate mail servers.

RIM could make this work because it controlled the hardware, the software, and the BlackBerry Enterprise Server middleware. Its rivals in those early days were the Palm Treo and Microsoft Windows Mobile. Palm was a bizarre beast that never really worked. Its owner, 3Com, first licensed the Palm OS to other manufacturers, then spun its software unit off into a separate company, PalmSource. The Treo was developed by one of those licensees, Handspring, which had been started by Palm’s founders. Palm eventually bought Handspring and reacquired some rights to the Palm OS, but it never had full control of the software. That’s a major reason why Palm and PalmOS gradually became non-competitive.

Microsoft’s mistakes were different, but illustrative of the traps inherent in an open phone operating system. In the best Windows tradition, Microsoft gave its handset manufacturers a lot of design freedom. It ended up with phones with a variety of screen sizes and configurations, with and without touchscreens, with and without physical keyboards. The hodgepodge of hardware made it impossible for Microsoft to provide a consistent–or particularly good–user experience on all Windows Mobile devices. And third-party software developers had a very hard time writing applctions that worked well, or sometimes at all, on all devices. In a final irony, until almost the very end, BlackBerry did a much better job of providing mobile access to Microsoft Exchange servers than Windows Mobile did.

Apple, of course, changed the game completely with the 2007 introduction of the iPhone, and again in 2010 with the iPad. Apple controls every aspect of the ecosystem, Apple software running on Apple hardware that can load only Apple-approved applications. This has horrified fans of open systems. such a Cory Doctorow and Jonathan Zittrain, but the mass market’s love for these devices has allowed Apple to suck up the lion’s share of profits in the handset industry and to define the tablet market to the point where it has no effective competition.

Except for Android, the open model has now all but collapsed. Nokia never achieved widespread adoption of Symbian by other manufacturers. Linux-based LiMo went nowhere, as did Nokia’s Maemo, Intel’s Moblin, and their love child, MeeMo.

The status of Windows Phone is uncertain. After the Windows Mobile nightmare, Microsoft set very tight design standards for its attempt to rejuvenate the platform. OEMs have a limited choice of display size and a physical keyboard is optional, but other specs must comply with the reference design. And Microsoft’s tight alliance with Nokia could result in, effectively, a line of “official” Nokia-built Windows Phone products. It’s nominally still a market where Microsoft offers its OS to any willing license, buy Redmond really controls the game.

Android’s openness has been a blessing and a curse. The free-to-all-comers OS has allowed the platform to gain a great deal of market share very quickly. It has also proved extremely frustrating to consumers, with a proliferation of designs and software versions all with different capabilities and no consistency in their ability to run third-party apps. With an iPhone, you know you will always be able to run the most recent version of the iOS software and any product in the App Store (with minor exceptions for some older models that lack some hardware features of more recent ones.) With Android, you just never know.

I suspect this will change in significant ways as a result of the Motorola Mobility acquisition. Google is never going to become Apple, but I suspect that the Android market is going to look a lot more like Windows Phone does today, with Motorola playing an even more central role than Nokia will for Microsoft. This sort of hybrid of open software with an official hardware maker is novel and largely untested; Palm and Nokia both nibbled at it, but neither was a fair test.

However it turns out, however, it looks like any attempt to build smartphones on the PC model is over.